This invention relates to acoustic cavitation, but more specifically, to an apparatus to produce acoustic cavitation using a single transducer to subject an article or component thereof to microcavitation events generated in a liquid insonification medium.
Acoustic microcavitation, which is the inducement of micron or sub-micron size bubbles in a liquid or fluid medium that survive a few microseconds or less, is to be contrasted with ultrasonic, megasonic, and cyrogenic aerosol cleaning methods. Microcavitation has been used on a limited scale or conceived for use in microparticle or sub-nanometer particle detection in ultrapure liquids, submicron particle eviction from silicon wafers, deinking of recyclable paper, paint removal, surgical procedures, destructive and non-destruction testing and measuring, thin film processing applications, etc.
Previously, at least two transducers were required to initiate and maintain cavitation. Prior ACIM was induced using a low frequency, high intensity primary acoustic field and a higher frequency, low intensity coaxing acoustic field. To effect ACIM, the two fields were substantially simultaneously directed at a site of a workpiece or object. It was crucial that at least part of the high frequency acoustic waves in the fluid medium pass the desired ACIM site precisely when the tensile part of the low frequency waves was present at the site. In this arrangement, it sometimes became unwieldy to articulate two transducers of different frequencies to achieve the desired ACIM zone, stationary or moving, where the different acoustic fields were to be synchronized and collocated.
Therefore, a need has arisen to simplify ACIM apparatuses and techniques to make them more practical to apply to the various applications identified herein.
In general, cavitation is the formation of cavities or bubbles in a liquid where the ensuing bubble dynamics and energy concentration result in implosive collapse of bubbles that achieve unique and surprising results. In the design of mechanical systems, cavitation has known destructive effects and therefore, was avoided. Cavitation remains enigmatic today as it was when Lord Rayleigh first investigated cavitational erosion of propellers almost a century ago. Cavitation is a mature subject and an encylopedic collection of information on acoustic cavitation is compiled in “Acoustic Bubble” by Tim Leighton (1997). Hydrodynamic cavitation is discussed in “Cavitation and Multiphase Flow Phenomena” by Frederick Hammitt (1980). Whether induced acoustically or associated with hydrodynamic flows, the mechanics and effects of cavitation are essentially the same. Acoustic cavitation has was also exhaustively reviewed by Flynn (1964), Neppiras (1979), Apfel (1981) and Prosperetti (1986).
Consider, for example, a free bubble in the path of a sound wave. In response to the sound wave, the bubble expands and contracts, and the energy mechanically stored during expansion is released in a concentrated manner during implosive collapse of the bubble. Should the bubble grow to about two and a half times its nominal or equilibrium size during negative excursions of acoustic pressure, then during the following positive half cycle of pressure, its speed of collapse could become supersonic (Lauterborn, 1969) thereby releasing excess energy that catastrophically explodes the bubble. Such almost single cycle violent events are called transient or inertial cavitation, and may explain the energetic manifestations of cavitation which, among other things, are useful for surface erosion or particle eviction.
Unlike dramatic bubble growth within a single acoustic cycle seen in transient or inertial cavitation, there exists a more gradual process, termed rectified diffusion. Under favorable conditions, a small bubble exposed to a continuous sound wave tends to grow in size if rectified diffusion is dominant. According to Henry's law, for a gas soluble in liquid, the equilibrium concentration of dissolved gas in the liquid is directly proportional to the partial pressure of the gas above the liquid surface, the constant of proportionality being a function of temperature. When the bubble expands, the pressure extant at the bubble's interior falls and gas diffuses into the bubble from the surrounding liquid. When the bubble contracts, the pressure in the interior increases and the gas diffuses into the solution of the surrounding liquid. The area available for diffusion, however, is larger in the expansion mode than in the contraction mode. Consequently, there is a net diffusion of the gas into the bubble from the surrounding liquid over a complete cycle, which causes bubble growth due to rectified diffusion.
However, a bubble can grow only up to a critical size—to a resonance radius determined by the frequency of the impressed sound wave. For small amplitude oscillations, a bubble acts like a simple linear oscillator of mass equal to the virtual mass of a pulsating sphere, which is three times the mass of displaced fluid. Stiffness is primarily given by the internal pressure of the bubble times the ratio of specific heats. Surface tension effects are, however, significant for small bubbles. Following Minnaert (1933) and ignoring surface tension, there is a simple relation for the resonance radius of air bubbles in water:(Resonance radius in μm)×(insonification frequency in MHz)=3.2
This relation is valid within 5% even for a bubble radius of about 10 μm. Bubble response becomes increasingly vigorous at the resonance radius, and is limited by damping mechanisms in the bubble environment—e.g., viscous damping, acoustic radiation damping, and thermal damping. A post-resonance bubble may exhibit nonlinear modes of oscillations, or become transient if the applied acoustic pressure amplitude is adequately high.
The above discussion presupposes the presence of a free bubble in the path of a sound wave. Free bubbles, however, do not last long in a body of water. Larger ones are rapidly removed due to buoyancy and the smaller ones dissolve even in nearly saturated water. While a 10 μm air bubble rises in water at a terminal speed of 300 μm/s, it can survive for about five seconds before dissolving completely. Dissolution is driven essentially by the excess pressure inside the bubble due to the surface tension.
It is very difficult to cavitate clean liquids (Greenspan and Tschiegg, 1967). A pure liquid purged of particulate impurities and stored in a perfectly smooth container can attain its theoretical tensile strength before undergoing cavitation or fracture. Under ideal conditions, water can be as strong as aluminum. The tensile strength of water based on the homogeneous nucleation theory exceeds 1000 bars. In cavitation studies, tensile strength is often quoted in terms of negative pressures, and cavitation threshold is understood as the pressure amplitude at which the first occurrence of cavitation is detected. Observed strengths (thresholds) in practice, however, are very much lower, rarely exceeding a few bars for reasonably clean liquids. This is because there exist gas pockets within the liquid which provide the necessary seeding for cavitation to occur at lower pressures.
A gas or cavitation site is often stabilized in a crevice (Harvey et al., 1944), either in a container wall or on a fluid-borne particle. Incomplete wetting traps gas at the root of a sharp crevice, stabilizing it against dissolution. Unlike a free bubble, though, surface tension in this case acts on a meniscus which is concave towards the liquid. Over-pressuring the liquid for sufficient duration prior to insonification can force the meniscus further into the crevice thereby causing full wetting of the crevice, which then gives rise to increased cavitation thresholds.
Until recently most acoustically generated, cavitation employed for cleaning applications, primarily used standing waves generated in a bath of liquid in which objects to be cleaned were immersed. In such ultrasonic cleaners, acoustic frequencies used were typically between 20 kHz to 100 kHz. Some implementations used propagating pulse trains instead of standing waves to improve cleaning efficiency, to minimize hot spot damage, and to reduce power consumption. Even so, when these applications were extended to semiconductor applications, cavitation was deemed detrimental to the delicate wafer surfaces, which spawned the use of megasonic cleaning to avoid cavitation (e.g. U.S. Pat. No. 4,854,337 to Bunkenburg et al., 1989; U.S. Pat. No. 4,979,994 to Dussault et al., 1990; U.S. Pat. No. 5,247,954 to Grant et al., 1993; and U.S. Pat. No. 5,355,048 to Estes, 1994) thus teaching the use of frequencies in the range of high kilohertz or low megahertz (typically 1 MHz).
Such high frequencies were used because it was believed that cavitation does not occur at higher frequencies. Quoting from the recent book edited by Takeshi Hattori (1998) titled, “Ultraclean Surface Processing of Silicon Wafers—Secrets of VLSI Manufacturing;” “[w]hen the oscillation frequency is 1 MHz or above, cavitation no longer occurs.” It is precisely the supposed inability of generating cavitation at low megahertz frequencies that such high frequency acoustics were used in diagnostic ultrasound for medical imaging and fetal monitoring. As a further precaution to preclude bubble growth that may occur due to continuous wave insonification, diagnostic instruments deployed short pulses at low duty cycles, e.g., 1%, which incidentally also facilitates the pulse echo method of information collection essential for their function. Therefore, prior systems rely on using high frequency tone burst acoustics, such as 1 MHz, when the explicit objective is to avoid the occurrence of cavitation.
Microcavitation, i.e., the inducement of micron or sub-micron size bubbles in a liquid or fluid medium that survive a few microseconds or less, occurs if the pressure amplitude in the acoustic beam is significantly greater than a threshold value, and if appropriate cavitation nuclei are present. In the absence of cavitation nuclei, water-like liquids cannot be fractured or cavitated by pressure amplitudes of less than 1000 bars peak negative, the threshold for homogeneous nucleation of water at standard temperature and pressure (STP), which corresponds to an atomic or molecular size vacancy or cavity in the liquid bulk caused by thermal, stochastic density fluctuations. Stronger tensile pressures are needed to cavitate smaller bubbles or cavitation nuclei. A 60-atmosphere peak negative pressure wave, for example, might cavitate a 50-nanometer bubble nucleus.
Planar piezoelectric transducers cannot generate very high pressure amplitudes with moderate power inputs. With increased power, however, cavitation might occur on the surface of the transducer crystal itself which will cause destruction of the crystal. By using focused transducers, however, it is possible to achieve additional pressure amplification by virtue of the focusing action at a particular site. Even so, high intensity acoustic waves invariably become non-linear because of inherent properties of the propagation medium. The nonlinearity in shape manifests an enhanced compressive peak and reduced tensile peak of the wave pulse. Cavitation at a nucleation site cannot occur if the tensile part of the wave is not stronger than the threshold value. If the nonlinear pulse is reflected at a pressure release boundary, then phase reversal takes place and the compressive peak reflects as a tensile peak and vice versa.
Using reflected nonlinear waves, it becomes easier to bring about cavitation because now a stronger tensile peak is available. U.S. Pat. No. 5,523,058 to Umemura et al. obviates the need for using suitable reflecting structures to achieve enhance tensile peaks by using two resonant transducers—one driven at a fundamental frequency and the second driven at a second harmonic frequency, and then superposing them in proper phase relation between the fundamental driving frequency pulse wave and its second harmonic wave to obtain a resultant pulse with enhanced tensile peak and weakened compressive peak. This method of generation, like other methods of cavitation in the past, also relies on the availability of appropriate cavitation nuclei in the insonified medium. Without the presence of appropriate nuclei the tensile peak is ineffective in causing cavitation.
Although Umemura teaches that “the efficiency of cavitation generation depends on the relative phase relation between a fundamental wave and a second harmonic” wave and he is able to access smaller bubble sizes (half the resonant bubble size corresponding to the fundamental frequency), he still relies on the availability of appropriate bubbles or bubble bearing crevice structures in the liquid host to initiate cavitation.
Further, Umemura does not use too high frequencies at which cavitation ordinarily does not occur. It is known in the art that transducers generating high pressure amplitudes at high frequencies are technologically unfeasible (high frequency resonant crystals are necessarily thin and cannot support stresses needed for generating high pressures), and yet to generate cavitation at high acoustic frequencies, the pressure amplitudes necessary are excessive.
In attempting to clean effectively throughout a cleaning tank, Honda (U.S. Pat. No. 5,137,580, 1992) uses at the bottom of the tank a Langvin type resonator with two resonating segments, and drives them alternately at the two resonance frequencies for periods of up to several milliseconds, which are adequate to setup standing wave fields in the liquid. At the lower frequency, a standing wave field causes large bubble cavitation to populate at pressure antinodes to form bubble bands at specific levels in the tank. At higher resonance frequency, Honda supposes that these bubbles cavitate and collapse to cause some measure of cleaning, but more importantly, because the standing wave pattern is broken, the previously structured bubble bands move upwards due to buoyancy and radiation forces to bring about some cleaning.
Honda suggests that these large bubbles will break down at higher frequencies and fill the tank with smaller bubbles. In actuality, the higher frequency waves merely reflect off the larger bubbles. A given frequency cannot significantly affect larger bubbles not corresponding to the characteristic resonance size. When the low frequency is again switched on, these small bubbles nucleate large bubble cavitation whose fragments will serve a next sweep by the higher frequency. Most cleaning is expected to be done by the large bubble cavitation effervescing throughout the extent of the tank. Honda does not explicitly state the frequencies he is using but the Langevin sandwich type transducer and the kind and scale of cavitation he mentions leads one to believe that he must be using acoustics in the low kilohertz range, between 20 kHz to 60 kHz.
If Honda were to use only one frequency, he would obtain a banded structure in the tank, and once the bubbles are setup in their locations, no significant cavitation would be sustained and no further cleaning effect would ensue due to occurrence of bubble effervescence. While Honda also teaches farming effectively available bubble fields for cavitation between two frequencies, Murry, before Honda taught how to cultivate bubble fields starting from the smallest of bubbles that he suggests are available in the liquid. Murry (U.S. Pat. No. 3,614,069, 1971) in his patent “Multifrequency Ultrasonic Method and Apparatus for Improved Cavitation, Emulsification and Mixing” teaches that operating on the assumption there will always be some very small bubbles in the bulk medium, insonification starts with using continuous wave insonification of a very high frequency corresponding to which the supposed pre-existing small bubbles are resonant. Near resonant bubbles exposed to continuous acoustic stimulus will respond by growing due to rectified diffusion. To continue this bubble growth they will have to be insonated by progressively decreasing the drive frequency. This downshifting insonification is achieved by using broadband transducers, not resonant transducers.
As the bubbles grow by downshifted continuous wave insonification, Murry applies a low frequency intense field to cavitate these bubbles. He upshifts or upconverts this low frequency to high intensity field so as to capture and cavitationally collapse any slightly smaller bubbles that may exist, as not all bubbles grow uniformly and simultaneously to a given size. Murry, operating on the assumption that very small bubbles exist in the liquid, concentrates on cultivating appropriate size bubbles by continuous wave insonification. Such bubbles are gas-filled as a result of rectified diffusion, they are not vacuous or nearly empty. Implosion of gas-filled bubbles is less energetic because the collapse is cushioned by the cavity contents.
Starting from a few tiny seed bubbles whose existence is assumed, Murry cultivates bubble fields with bubbles progressively growing over time in response to frequency downshifted insonification, and then violently collapsing them by applying low frequency high intensity acoustic field, the latter being subsequently upshifted in frequency to harvest all possible bubbles for cavitation. He uses two broad-band transducers to facilitate frequency shifting, and even interchanges the roles of the bubble grower and bubble exploder transducers for appropriate cycling and sustaining cavitation throughout the extent of the bulk being processed for emulsification or mixing.
In summary the prior art teaches that a perfectly clean liquid absent of bubbles or bubble-like structures cannot be easily cavitated. To bring about cavitation in ultra clean hosts, especially at high frequencies, is almost impossible primarily because the acoustic drivers, the piezoelectric transducers used to generate cavitation cannot be made to generate high pressure amplitude sound waves at high frequencies. It is possible to a limited extent to generate high tensile pulses, but only with reduced compressive pulses if one drives the transducer in both fundamental and second harmonic excitation in precise phase relationship.
To achieve this, one must use two transducers. In resonant mode excitation, the transducer can only be driven at odd harmonics of the fundamental frequency. Even if one is able to obtain high pressure amplitude at high frequency, one needs to assume that a population of small bubbles always exist in a liquid, then insonifying the liquid medium with continuous acoustic waves of appropriately high frequency, frequency specific to excite resonance in the bubbles, can grow the bubbles to a larger size through rectified diffusion, whence subsequent insonification by a lower frequency of sufficient intensity one can bring about cavitation. Being gas-filled these long-lived bubbles cannot sufficiently implode to create high energy density points in the medium, and are thus ineffective to bring about the effects of ACIM described herein.
It is known in physics of liquids that free bubbles in a liquid are unstable and do not survive for any significant duration after their creation. Larger bubbles rise and escape out of the liquid because of buoyancy, while smaller bubbles dissolve due to surface tension forces which are dominant for small bubbles. Any bubble-like structure that survives in liquid has to be anchored in a crevice like feature in a solid, e.g., a wall or liquid-borne particle. Not all liquid borne particles are capable of supporting such partially wetted crevices, particularly, smooth spherical particles cannot harbor such gas-filled cavities.
Apart from the inventor's own work, the teachings of the entire prior art appears to rely on cavitation as a chance dominated phenomenon. In addition, it is not taught or suggested in the prior art how to create cavitation nuclei when none exist a priori, and then to control such cavitation after onset.
Therefore, to achieve useful applications provided by the present invention in a practical and convenient manner, prior systems and methods do not take into account: (i) how to activate or nucleate a cavitation event from a particle, regardless of whether or not it has a gas bearing crevice, (ii) how to acoustically activate or nucleate cavitation amongst particles, however, small they may be, or whatever be their composition or surface morphology, (iii) consideration of the number of times a cavitation event ensues in relation to a given or created gas bearing crevice and/or point phase boundary, or (iv) attaining vacuous cavitation to the maximum extent possible rather than gaseous cavitation.
In vacuous cavitation the cavity is nearly empty. Only transiently (or inertially) generated cavitation involves vacuous cavities. Cavitation generated by continuous waves is gaseous cavitation. Only vacuous cavitation can be imploded, unimpeded, unto a point, and hence, only vacuous cavitation can culminate in high energy density at points. To be able to implement items (i) through (iv) implies that cavitation is being constructively controlled in all phases—inception, evolution and intensity.
To the inventor's knowledge, the entire prior art concerns itself with cavitation as chance dominated phenomenon, and does not teach how to manage cavitation in a practical and efficient way to perform a useful purpose, except in the inventor's two recent U.S. Pat. Nos. 5,681,396 (1997) and 5,594,165 (1997), which deal with acoustic coaxing methods for constructive control of the cavitation phenomenon using confocal transducers.
ACIM methods described herein, on the other hand, employ a single transducer to more effectively control the onset, evolution and intensity of microcavitation. Generating ACIM with a single transducer enables expanded utility including, improved deinking of paper (e.g., removal of bonded, laser printed Xerox ink, i.e., toner-based ink compositions), practical depainting of surfaces (including selective removal of layers in a multi-layered painted surface (primer and/or top coat)), thin film strength testing and surface preparation prior to thin film deposition; semiconductor wafer cleaning; improved microparticle detection in clean liquids; improved particle removal for precision cleaning of delicate surfaces; and better particle size control in the preparation of nanometer particles like gold sols. In addition, improved ACIM methods and apparatuses of the present invention may be used to erode metallic surfaces, help shatter kidney stones, accelerate chemical reactions and even lead to light production, i.e., sonoluminescence.